Test-infra: Make plank an actual Kubernetes controller

[alvaroaleman]

What would you like to be added:

Plank to be changed to be an actual Kubernetes controller

Why is this needed:

Plank is currently implemented as a CronJob with a 30 seconds interval. It would be great if it was an actual controller instead that gets notified whenever a ProwJob or Pod resource changes, because that would allow it to instantly react which would be a nicer user experience.

Also, having a sync loop that acts on one ProwJob rather than a huge list of ProwJobs will most likely simplify the code and make it more efficient, as we will stop looping over outdated ProwJobs CRs.

/area prow
/area prow/plank

[alvaroaleman]

I would imagine the whole transition to roughly go like this:

• Plank gets updated to use Listers for ProwJobs
• Plank gets updated to use Listers for Pods
• A featuregated "PlankV2" implementation gets added that is an actual controller. All existing tests get executed and pass against it
• Interested parties try out PlankV2 and provide feedback
• We start being confident in PlankV2, make it the default and delete the current implementation

[stevekuznetsov]

In a perfect world this happens after GitHub work is removed from plank and resides squarely in sinker. @krzyzacy what's the final status of that?

[alvaroaleman]

@stevekuznetsov you are talking about the reporting and mean [...] and resides squarely in crier or not?

The "Plank reports to GitHub" part is deprecated and redundant with what crier does, so we can just not add it to a PlankV2

[stevekuznetsov]

Yep, agreed, but IIRC we've had some small issues with rollout and OpenShift's deployment is not on crier yet so I'd hope we all migrate to using it before plank-2:1.0.0 is ready

[krzyzacy]

in a perfect world this happens after GitHub work is removed from plank

I have no idea with the status for all the prow deployments in wild, but at least our deployments are on crier now.

[stevekuznetsov]

Sweet, and we have fixed all the weird PATCH issues? Is there any concise doc on how to switch over?

[alvaroaleman]

@stevekuznetsov It does PATCH and we have apparently multiple redundant fixes, namely both the crier controller doing wait.Poll to make sure it doesn't act on an outdated cache _and_ hardcoding the GitHub reporter to max one worker.

And it seems we are also using Plank for GitHub and crier only for Slack reporting

[stevekuznetsov]

I am ready to move once someone shows me how!

[krzyzacy]

I still need to get https://github.com/kubernetes/test-infra/pull/13952 in... (It's hard to work on OSS and internal at the same time, different dev envs, but let me rebase that PR)

For doc, there is https://github.com/kubernetes/test-infra/tree/master/prow/crier#migration-from-plank-for-github-report, I'm not too up to date with all the controllers, I can read it again.

[alvaroaleman]

/assign

[fejta-bot]

Issues go stale after 90d of inactivity.
Mark the issue as fresh with /remove-lifecycle stale.
Stale issues rot after an additional 30d of inactivity and eventually close.

If this issue is safe to close now please do so with /close.

Send feedback to sig-testing, kubernetes/test-infra and/or fejta.
/lifecycle stale

[fejta-bot]

Stale issues rot after 30d of inactivity.
Mark the issue as fresh with /remove-lifecycle rotten.
Rotten issues close after an additional 30d of inactivity.

If this issue is safe to close now please do so with /close.

Send feedback to sig-testing, kubernetes/test-infra and/or fejta.
/lifecycle rotten

[fejta-bot]

Rotten issues close after 30d of inactivity.
Reopen the issue with /reopen.
Mark the issue as fresh with /remove-lifecycle rotten.

Send feedback to sig-testing, kubernetes/test-infra and/or fejta.
/close

[k8s-ci-robot]

@fejta-bot: Closing this issue.

In response to this:

Rotten issues close after 30d of inactivity.
Reopen the issue with /reopen.
Mark the issue as fresh with /remove-lifecycle rotten.

Send feedback to sig-testing, kubernetes/test-infra and/or fejta.
/close

Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes/test-infra repository.

[alvaroaleman]

/reopen

[k8s-ci-robot]

@alvaroaleman: Reopened this issue.

In response to this:

/reopen

Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes/test-infra repository.

[fejta-bot]

Rotten issues close after 30d of inactivity.
Reopen the issue with /reopen.
Mark the issue as fresh with /remove-lifecycle rotten.

Send feedback to sig-testing, kubernetes/test-infra and/or fejta.
/close

[k8s-ci-robot]

@fejta-bot: Closing this issue.

In response to this:

Rotten issues close after 30d of inactivity.
Reopen the issue with /reopen.
Mark the issue as fresh with /remove-lifecycle rotten.

Send feedback to sig-testing, kubernetes/test-infra and/or fejta.
/close

Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes/test-infra repository.

[alvaroaleman]

/reopen
This is actively being worked on in https://github.com/kubernetes/test-infra/pull/17626

[k8s-ci-robot]

@alvaroaleman: Reopened this issue.

In response to this:

/reopen
This is actively being worked on in https://github.com/kubernetes/test-infra/pull/17626

Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes/test-infra repository.